This invention relates to electric motors, and more particularly, to apparatus for removing heat from high-power electric motors, in which at least a portion of the heat removal system is integrated in a housing of the motor, and to methods for constructing such motors. A variety of electric motors have long been used in industrial applications, and many techniques have been developed to remove the heat produced by such motors. However, a class of electric motors, which are often referred to as "Brushless DC" motors, have recently become the preferred motor for use in many electrically-powered industrial drive systems. Although motors of this class provide several advantages over other motor types, they also present special heat removal challenges.
A significant advantage of brushless DC motors over many other types of motors is that the motor's speed can be accurately variably controlled over a wide range of speeds and load conditions, throughout which the motor can produce full, or nearly full rated torque. For example, a typical brushless DC motor may produce 90 percent of its maximum rated torque at speeds as low as five percent of the motor's maximum speed. Although variable-speed motors have long been known, prior variable-speed motor technologies have generally not produced nearly full rated torque at small fractions of a motor's maximum speed. For typical motors, output torque, and other measures of useful output, drop dramatically as speed is reduced. As a result, when an application required a variable speed motor to produce high torque at low speed, a motor capable of greater output than needed at normal speeds was often selected in order to provide sufficient output at very low speeds.
Although the practice of selecting an oversized motor for such applications is generally undesirable, it does provide advantages in coping with the waste heat produced during motor operation. An oversized motor is typically operated at a smaller fraction of its maximum output or duty cycle, thereby reducing its heat production in comparison to what was contemplated by its designers for full-load operation. In addition, higher capacity motors are physically large. This limits the number of motors which may be employed in machinery of a given volume. Also, larger motors have greater internal mass and external surface area over which heat produced in the motor may be distributed. The conventional method of cooling medium-to-large electric motors used in industrial equipment is to direct a forced air stream at the exterior surface of the motor. The rate at which heat is removed from the motor depends in large part on the surface area available for contact with the air stream.
The availability of variable-speed DC brushless motors capable of high output across a wide speed range allows equipment designers to specify smaller capacity motors compared to those required when other technologies are used. Such applications present special heat-removal challenges. Because the motors are physically smaller, all the heat produced must be removed from a smaller surface area. Also, more motors can fit into a given volume within an equipment enclosure. Further, when lower-capacity motors are used, they must operate at a higher fraction of their full-load output or maximum duty cycle. As a result, waste heat must be removed efficiently to avoid an excessive temperature increase which could damage the motor or the equipment in which it is installed, and could cause a fire or other safety hazard.
As noted above, a preferred conventional method of cooling medium-to-large electric motors used in common industrial equipment is to direct a forced air stream at the exterior surface of the motor. Although this method may be effective, it has several disadvantages. Since the air stream must actually contact the motor exterior surface, substantial space must be provided around the motor to permit rapid air movement. Normally, a blower or fan is used to move the cooling air. Because DC brushless motors may be required to operate at low speed, the motor cannot drive its own blower, and therefore, a smaller but not-insignificant additional motor is typically provided to drive the blower or fan blade. The additional motor requires energy and space and generates still more heat. Since air is not an efficient coolant in comparison with other cooling materials, the cooling air must strike the motor surface and circulate around it at relatively high velocity.
As a result, in order to obtain the desired velocity of air, high-velocity blowers must be used, and these blowers must be placed relatively close to the motor to be cooled. A substantial amount of space must be provided around each motor to allow enough air circulation. These consequences of using forced air cooling limit the density of the equipment in the region of the motors. The blowers produce high-intensity noise, which may be hazardous to human workers in the vicinity. In addition, the blowers have rapidly moving mechanical components, which may also be hazardous to workers. Thus, when blowers are used, special care must be taken to comply with occupational safety and health regulations.
In addition, the cooling air must be drawn from and returned to the environment surrounding the equipment. Industrial facilities are rarely perfectly clean, and thus, the environmental cooling air often carries particulates which may contaminate the product being manufactured or processed, or the equipment itself. Although filtration of the air may be attempted, such filtration may not be completely effective. Further, there are some environments in which blowers may be especially hazardous. For example, in some mining and materials processing operations, an explosive atmosphere may be caused by an explosive gas or by suspended particles in the air which may ignite explode, and rapid air movement may be undesirable.
It is generally desirable to operate motors at low internal and external temperatures (e.g., approximately room temperature), to minimize the risk of mechanical and electrical failures which are promoted by higher temperatures. However, it is difficult in practice to accomplish this. Although improved cooling performance can generally be obtained by increasing the volume and velocity of the circulating air, the initial cost, operating cost, space requirements, and other factors associated with larger air movement equipment limits the additional cooling which can be feasibly achieved in this way. Thus, even with forced air cooling, the exterior temperature of the housing of a typical motor in industrial equipment may be high enough to cause a burn if the housing is touched.
Another disadvantage of using blowers to cool motors is that the heat produced within the motor must be conducted to the outermost surface of the motor housing before it may be transferred to the cooling air. Conventionally, blower-cooled motors have been provided with a plurality of cooling fins to increase their surface area, thereby providing improved thermal conductivity between the housing and the cooling air. The additional fin structure increases the mass, material cost, and space requirements of the motor. Further, although the materials (such as aluminum) typically used to form motor housings are thermally conductive, they are not perfectly conductive, and present some thermal resistance. To avoid undesirably high motor operating temperatures in general, and to avoid "hot spots" in particular, the exterior geometries selected for housings of blower-cooled motors is limited to those which provide sufficient total cooling, and those which are sufficiently uniform to avoid localized "hot spots." As a result, most housings for air-cooled motors are approximately cylindrical, even though other exterior shapes would be preferred if cooling were not an issue.
Several other methods have also been used to cool motors, but these too have their disadvantages. According to one known cooling method, a cooling jacket is attached to the exterior of the motor housing. The cooling jacket has one or more circuitous passages through which a liquid coolant flows. The circuitous passages are formed between a pair of sheet members which are joined in various locations to create interstitial passages and are cylindrically constructed to substantially conform to the shape of the housing.
Because the jacket is constructed of plate members which must be joined together to form the interstitial passages, the manufacture of the jacket is fairly complex. In addition, it is difficult to permanently prevent leakage throughout the intended operating life of the motor. Since coolant liquids may be conductive and may chemically attack certain materials, even minute leakage of coolant in or around the motor may result in a catastrophic failure of the motor and may cause a personal safety hazard. Also, the jacket is, in essence, a separate assembly which must be installed on the exterior of the motor housing. This separate installation increases cost and complexity.
Further, it is difficult to install the jacket on the exterior of the motor housing such that the housing and the jacket are in continuous contact along their entire adjacent surfaces, thereby forming a highly-thermally-conductive interface therebetween. Voids between the housing and the jacket reduce the overall cooling efficiency, and create localized regions of high temperature (i.e., "hot spots") which may cause the failure of insulation or other motor components. In addition, the range of available housing shapes is limited to those for which a conformal jacket may be produced and subsequently conveniently installed on the housing. Thus, for motor housings having an exterior cross-section which is not exactly circular, the jacket may be unusable. For most medium to large motors, although the housing cross section is generally circular, the circular shape is typically modified to provide mounting feet or other support structures, or to provide attachment surfaces for control equipment and the like. Thus, in many motor applications, the cooling jacket of the prior art may not be advantageously applied.
Another known method of cooling electric motors and other rotating electrical machines (such as generators) is to form passages for carrying a cooling fluid on an interior surface of the motor housing itself. This has typically been accomplished by casting, machining, or stamping a suitable pattern of indentations or channels on the interior structural wall of the housing. A cooperating closure, such as a suitably shaped plate member, is placed over the indentations or channels to complete the passages and prevent the fluid from escaping. This method also has several disadvantages which prevent it from being usefully applied to medium to large scale DC brushless motors.
One problem is that it is difficult to create a leak-proof seal between the closure and the housing. Motors are subject to vibration and large mechanical stresses which cause some structural deflection during operation. In at least one prior art device, the housing is essentially cylindrical, and the mating closure member is also cylindrical and is telescopically arranged therein. Such mechanical arrangements require that the closure and housing be sealed at each end of the motor over the entire circumference of the closure-to-housing interface. For motors of medium to large size, a seal along such a large circumference which remains permanently reliable despite vibration, mechanical deflection of the housing, and many thermal expansion and contraction cycles, is difficult to form. Further, the cooling fluid channels of prior art motors and the closure plate member covering the interior surface of the housing are believed to degrade the structural integrity of the housing.
In addition, it is believed that providing an interior closure member to be applied to the inside of the motor housing is incompatible with an aspect of the modern preferred techniques of constructing DC brushless motors. Typically, housings of such motors are formed with an inside diameter which is slightly smaller than the outside diameter of the stator winding assembly at normal operating temperatures. The stator winding assembly is permanently secured in the motor housing by heating the housing well above normal operating temperatures to cause thermal expansion, inserting the stator winding assembly into the proper position in the housing cavity, and allowing the housing to cool. As the housing cools, it contracts around the stator winding assembly to produce a secure "compression-fit" engagement between the housing and the winding. This engagement is essential because all of the motor's output torque is transferred from the stator winding assembly to the housing at this interface. This construction method complicates the problem of a seal between an interior closure and the motor housing, because the large forces produced by the stator winding-to-motor housing compression fit would tend to deform the closure. Variable torques applied at this interface during motor operation further exacerbate this problem.
In addition, the cooling fluid passages may be difficult to form on the interior wall surface of a complete cylinder. Accordingly, in at least one case, the motor housing has been formed from two mating semi-cylindrical components with suitable mating plate-like closures. The semi-cylindrical housing components are later joined in a clam-shell arrangement. It is believed that this construction is particularly disadvantageous in DC brushless motor applications because another means must be provided to permanently secure the stator winding assembly to the housing components.